In normal caregiver-infant dyads, caregivers regulate infant attention and state of arousal and affect by closely monitoring the infant and providing appropriate vocal, facial, gestural, kinesthetic and tactile stimulation. Mothers with symptoms of depression fail to provide needed emotional and state regulation for their infants, who exhibit disordered state and behavior in the dyad and who are at risk for later problems of emotion, attachment, and cognition. A substantial body of research shows that caregivers use the vocal channel in particular to modulate infant attention, affect, arousal, and information-processing. Infant-directed (ID) speech possesses exaggerated modulation of the fundamental frequency (Fo) relative to adult-directed (AD) speech, and is known to elicit stronger behavioral and affective responding from infants. Research on the ID speech produced by mothers with symptoms of depression has shown that it tends to have less exaggerated Fo modulation. In a preliminary study involving an associative learning paradigm in which 4-month-olds of nondepressed mothers were taught an association between a voice and a smiling face, ID speech segments from mothers with relatively few symptoms of depression were readily associated with the face, whereas those segments produced by mothers with comparatively more symptoms of depression failed to enter into association with the face. The ID speech of mothers with more symptoms of depression lacked the attention-getting rising Fo contours that are characteristic of normal ID speech. The proposed experiments will seek to replicate this finding, extend the study to mothers with more severe symptoms of depression, and explore in greater detail the acoustic correlates of ID speech produced by these mothers. In addition, associative learning in infants of depressed mothers will be studied to determine if these mothers can associate the ID speech of their mothers, that of other depressed mothers, or that of nondepressed controls with face stimuli. The concurrent and predictive validity of associative learning for performance on the Bayley Scales of Infant Development will also be assessed. The importance of the proposed research lies in its promise to elucidate one mechanism by which maternal depression influences infant associative learning, thereby suggesting strategies to reduce risk for cognitive deficits.